A GPS-based framework for accelerated vehicle-level durability evaluation

Abstract

Accelerating durability testing has become a pivotal requirement in modern vehicle development due to increasingly compressed time-to-market demands and stringent cost targets. Accelerated Durability Testing (ADT) provides an essential framework for reproducing years of cumulative service damage within a shortened testing period while maintaining a strong correlation with real-world conditions. However, conventional ADT strategies remain largely empirical, constrained by fixed acceleration factors, limited statistical grounding and insufficient integration of field data and physics-based methods. Moreover, although state-of-the-art ADT methodologies enhance acceleration frameworks by integrating load reconstruction, damage equivalence and digital validation techniques, they still fail to holistically bridge the gap between consumer usage data and systematic test design, and they rely heavily on costly hardware infrastructure. To address these issues, this study introduces a novel data-driven methodology for synthesising vehicle-level durability evaluation frameworks by combining statistical clustering with GPS-derived effective load values and incorporating load correlation across subsystems. The proposed approach systematically segments real-world routes and Micro-Trips, using approximately 157,000 km of GPS-based fleet driving data from 84 drivers across Iran and considering daily traffic, route topography and design margins, to construct durability evaluation cycles that are both time-efficient and representative of customer usage. The synthesised ADT cycle is validated against a statistically representative 1600 km target cycle through descriptive statistical metrics. The cumulative damage results for the gearbox and differential systems confirm that the accelerated cycle preserves the damage equivalence of the target spectrum while achieving a 5.6-fold reduction in reduction in equivalent durability mileage. The proposed ADT is formulated as a damage-equivalent durability evaluation spectrum that can be mapped onto proving grounds rather than as a literal public-road test route. These findings demonstrate that the framework integrates real-world usage characteristics with structured durability evaluation, enabling scientifically grounded, cost-efficient and physically interpretable ADTs for modern automotive validation programmes.